
# Importing Necessary Libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from collections import Counter


import plotly.express as px
from wordcloud import WordCloud

import re
import nltk
from nltk.corpus import stopwords
import warnings
warnings.filterwarnings("ignore")


# Import the CSV file, save it in a data frame, and display the top 3 rows
df = pd.read_csv('content.csv')
df.head(3)


print('Total no. of rows:',df.shape[0])
print('Total no. of columns:',df.shape[1])

Total no. of rows: 8807
Total no. of columns: 12


# info about the columns and its data types 
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8807 entries, 0 to 8806
Data columns (total 12 columns):
 #   Column        Non-Null Count  Dtype 
---  ------        --------------  ----- 
 0   show_id       8807 non-null   object
 1   type          8807 non-null   object
 2   title         8807 non-null   object
 3   director      6173 non-null   object
 4   cast          7982 non-null   object
 5   country       7976 non-null   object
 6   date_added    8797 non-null   object
 7   release_year  8807 non-null   int64 
 8   rating        8803 non-null   object
 9   duration      8804 non-null   object
 10  listed_in     8807 non-null   object
 11  description   8807 non-null   object
dtypes: int64(1), object(11)
memory usage: 825.8+ KB


# change the datatype for date added to datetime for better analysis
df['date_added'] = df['date_added'].apply(pd.to_datetime)


df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8807 entries, 0 to 8806
Data columns (total 12 columns):
 #   Column        Non-Null Count  Dtype         
---  ------        --------------  -----         
 0   show_id       8807 non-null   object        
 1   type          8807 non-null   object        
 2   title         8807 non-null   object        
 3   director      6173 non-null   object        
 4   cast          7982 non-null   object        
 5   country       7976 non-null   object        
 6   date_added    8797 non-null   datetime64[ns]
 7   release_year  8807 non-null   int64         
 8   rating        8803 non-null   object        
 9   duration      8804 non-null   object        
 10  listed_in     8807 non-null   object        
 11  description   8807 non-null   object        
dtypes: datetime64[ns](1), int64(1), object(10)
memory usage: 825.8+ KB


df.head(3)


# putting the column names in a list
column_names = df.columns.tolist()
null_count = df.isnull().sum()
null_ratio = round(df.isnull().mean(),3)
nunique = df.nunique()
#display the dataframe with all the information
pd.DataFrame(np.vstack((column_names,null_count,null_ratio, nunique)).T, 
             columns = ['Columns','Null Count','Null Ratio','Unique Values'])


# checking presence of duplicate records
df.duplicated().sum()

0


df['rating'].unique()

array(['PG-13', 'TV-MA', 'PG', 'TV-14', 'TV-PG', 'TV-Y', 'TV-Y7', 'R',
       'TV-G', 'G', 'NC-17', '74 min', '84 min', '66 min', 'NR', nan,
       'TV-Y7-FV', 'UR'], dtype=object)


invalid_rating = df[df['rating'].isin(['74 min', '84 min', '66 min'])]
invalid_rating


for i,r in invalid_rating.iterrows():
    df['duration'][i] = r.loc['rating']
    df['rating'][i] = np.nan
df['rating'].unique()

array(['PG-13', 'TV-MA', 'PG', 'TV-14', 'TV-PG', 'TV-Y', 'TV-Y7', 'R',
       'TV-G', 'G', 'NC-17', nan, 'NR', 'TV-Y7-FV', 'UR'], dtype=object)


#check validity in numerical data
df.describe()


#Distribution of Type of Contents
type_dist = df['type'].value_counts()
type_dist

Movie      6131
TV Show    2676
Name: type, dtype: int64


#visualizing the distribution in a barplot
fig = plt.figure(figsize=(6,5)) #setting the figure size
sns.barplot(x = type_dist.index,
                       y = type_dist.values, 
                       palette = 'crest')
plt.xlabel('Type of contents')
plt.ylabel('Total number of contents')
plt.title('Distribution of content type')
plt.show()


# Top 20 directors for the contents
top20_dirs = df['director'].value_counts().head(20)
top20_dirs

Rajiv Chilaka             19
Raúl Campos, Jan Suter    18
Marcus Raboy              16
Suhas Kadav               16
Jay Karas                 14
Cathy Garcia-Molina       13
Martin Scorsese           12
Youssef Chahine           12
Jay Chapman               12
Steven Spielberg          11
Don Michael Paul          10
David Dhawan               9
Yılmaz Erdoğan             8
Lance Bangs                8
Kunle Afolayan             8
Quentin Tarantino          8
Ryan Polito                8
Troy Miller                8
Shannon Hartman            8
Fernando Ayllón            8
Name: director, dtype: int64


#visualizing the top 20 directors in a barplot
fig = plt.figure(figsize=(6,5)) #setting the figure size
sns.barplot(x = top20_dirs.values,
            y = top20_dirs.index, 
            palette = 'crest',
            )
plt.xlabel('Total number of contents directed')
plt.ylabel('Directors')
plt.title('Top 20 directors')
plt.xticks(range(0,max(top20_dirs)+2,2))

plt.show()


#printing values of cast for top 5 rows
df['cast'].head()

0                                                  NaN
1    Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban...
2    Sami Bouajila, Tracy Gotoas, Samuel Jouy, Nabi...
3                                                  NaN
4    Mayur More, Jitendra Kumar, Ranjan Raj, Alam K...
Name: cast, dtype: object


def combine_and_count_values(data, data_name, top = None):
    # Parameters --> data: column of dataframe, 
    #                data_name: string for name of data, 
    #                top: int if top n numbers needed
    # Returns a dataframe with sorted values by its count

    
    ##combine all the non null values from the data
    data = [d for d in data if not pd.isna(d)]

    #join all the strings into one
    collection = ','.join(data)
    
    #separate them to a list
    list_collection = [c.strip() for c in collection.split(',')]

    #use counter library to get the counts and put them in dataframe
    counter = Counter(list_collection)

    #transpose to get desired output
    df = pd.DataFrame([list(counter.keys()),list(counter.values())]).T
    #setting the name of columns
    df.columns = [data_name,'Count']
    
    if top:
        return df.sort_values('Count', ascending = False, ignore_index = True).head(top) #returns top n elements
    else:
        return df.sort_values('Count', ascending = False,ignore_index = True) # returns all elements


top20_actors = combine_and_count_values(df['cast'], 'Cast Member', top=20)
top20_actors


#visualizing the top 20 actors in a barplot
fig = plt.figure(figsize=(6,5)) #setting the figure size
sns.barplot(x = top20_actors['Count'],
            y = top20_actors['Cast Member'], 
            palette = 'crest',
            )
plt.xlabel('Total number of appearances')
plt.ylabel('Actors')
plt.title('Top 20 Most Appearing Actors')
plt.xticks(range(0,max(top20_actors.Count)+2,2))

plt.show()


## Distribution of country
df['country'].value_counts()

United States                             2818
India                                      972
United Kingdom                             419
Japan                                      245
South Korea                                199
                                          ... 
Romania, Bulgaria, Hungary                   1
Uruguay, Guatemala                           1
France, Senegal, Belgium                     1
Mexico, United States, Spain, Colombia       1
United Arab Emirates, Jordan                 1
Name: country, Length: 748, dtype: int64


country_data = combine_and_count_values(df['country'], 'Country')
country_data


import math
def create_map(country_data):
    country_data['log_count'] = [math.log(count) for count in country_data['Count']]
    fig = px.choropleth(country_data, locations='Country',
                        locationmode='country names',
                        color= 'log_count', #use log to color as the difference between counts are significant
                        color_continuous_scale= px.colors.sequential.Hot_r,
                        custom_data=['Country', 'Count']
                       )
    
    fig.update_traces(hovertemplate='<b>%{customdata[0]}</b><br>Count: %{customdata[1]}')
    # update layout figure method
    fig.update_layout(
        title_text='',
        width=1200,
        height=600,
    )
    
    fig.show()
create_map(country_data)


top_10_countries = country_data[['Country','Count']].head(10)
others = pd.DataFrame([['Others',np.sum(country_data[10:]['Count'])]], columns = ['Country','Count'])
top_countries = pd.concat((top_10_countries,others), axis=0)
top_countries


fig = px.pie(top_countries, values='Count', names='Country',         
             hover_data=['Count'], 
             labels={'Country':'Country'},
             title= 'Percentage of contents created per countries',
             height=600,
             width=1200)
fig.update_traces(textposition='outside', textinfo='percent+label')
fig.update_layout(title={'x':0.5, 'xanchor': 'center'})
fig.show()


def category_per_country(cat):
    # splitting the data with , as some have multiple values and convert them in a list
    c_df = df['country'].apply(lambda x: [s.strip() for s in x.split(',')] if isinstance(x,str) else x)
    
    # explode creates a new record for each of the data in the list and to frame is to convert it into dataframe
    c_df = c_df.explode().to_frame()

    # create a new column cat where each country gets its value as per the index form original dataframe
    c_df[cat] = c_df.apply(lambda x:df[cat][x.index])

    # create a final table to create dummy variable for each value in table and calculate the count of each cat
    final = c_df.pivot_table(index='country', columns=cat, aggfunc='size', fill_value=0)

    #temp total column to sort the table
    final['total'] = final.sum(axis = 1)
    return final.sort_values('total', ascending = False).drop(columns =['total'])

type_dist = category_per_country('type')
top_10_countries = type_dist.head(10)
top_10_countries.reset_index(inplace = True)
top_10_countries


# top_10_countries.plot(x='country'.index, y=["Movie", "TV Show"], kind="bar") 
fig, ax1 = plt.subplots(figsize=(10, 6))
tidy = top_10_countries.melt(id_vars='country').rename(columns=str.title)
sns.barplot(y='Country', x='Value', hue='Type', data=tidy, ax=ax1, palette= 'crest',)
ax1.set_xlabel('Number of contents')
ax1.set_title('Type of contents created per country')
sns.despine(fig)
plt.show()


#creating rating column
rating_dist = category_per_country('rating')
rating_dist


for i,country in rating_dist.head(10).iterrows():
    fig = px.pie(country.index, values=country.values,names = country.index, 
                 labels={'Country':'Country'},
                 title= 'Percentage of contents created per countries',
                 height=600,
                 width=1200)
    fig.update_traces(textposition='outside', textinfo='percent+label')
    fig.update_layout(title={'x':0.5, 'xanchor': 'center'})
    fig.show()


def genere_per_country(data,count_list,top_number):
    """
    arg: data --> gets the dataframe to check and unstack the generes present in the data
    arg: count_list --> column which needs to be calculated
    arg: top_number--> most values required
    func: explode the genre column and get thhe top 10 countries most popular generes overall 
    return: data table for plot, top counts of the column
    """
    data=data.copy()
    new_column='new_{}'.format(count_list)
    
    data[new_column]=data[count_list].str.split(', ')
    temp_explode=data.explode(new_column)
    top_countries_content=temp_explode['country'].value_counts().head(10).index
    temp_explode_1=temp_explode[temp_explode['country'].isin(top_countries_content)]
    top_new_col_count=temp_explode_1[new_column].value_counts().head(top_number).index
    temp_explode_2=temp_explode_1[temp_explode_1[new_column].isin(top_new_col_count)]
    final_table_group=temp_explode_2.groupby(['country',new_column]).size().unstack()
    count_table = temp_explode_2.groupby(['country', new_column]).size().reset_index(name='count')
    final_table_group=final_table_group.fillna(0).astype(int)
    
    return top_new_col_count,final_table_group,count_table


#calling the function
top_genere,grp_df,count_table=genere_per_country(df,'listed_in',5)


#top genre based on the cumulative counts for a country 
top_genere

Index(['Dramas', 'International Movies', 'Comedies', 'International TV Shows',
       'Documentaries'],
      dtype='object')


#grouped table on the top 10 countries and their top genre
grp_df


## visualization of the above sorted table


plt.figure(figsize=(10,6))
sns.barplot(y='count',x='country',hue='new_listed_in',data=count_table)
plt.xlabel('countries')
plt.ylabel('values')
plt.title('genere-content across countries')
plt.tight_layout()
plt.xticks(rotation=90)
plt.show()


va=['top','bottom']
color = ['red','green']
margin = [-20,20]
plt.figure(figsize=(10,8))
for t in df['type'].unique():
    a = df[df['type']==t]
    #grouping by just the year of the date added
    year_group = a.groupby(df['date_added'].map(lambda x: x.year))['show_id'].count()
    color_ = color.pop()
    va_ = va.pop()
    margin_ = margin.pop()
    sns.lineplot(x = year_group.index,
                 y = year_group.values,
                 label = t,
                 color = color_, 
                 marker = 'o')
    
    for x, y in zip(year_group.index,year_group.values):
        plt.text(x,
                 y+margin_,
                 y,
                 horizontalalignment='center',
                 verticalalignment=va_,
                 bbox=dict(facecolor=color_, alpha = 0.5)) #some decorations
# Add labels and title
plt.title('Type of content added over time')
plt.legend(title='Type', bbox_to_anchor=(1, 1), loc='upper left')
plt.xlabel('Year')
plt.ylabel('Count')
plt.show()


from nltk.corpus import stopwords


stop_words = stopwords.words('english')


titles = df['title']
# Create and generate a word cloud image:
title_wordcloud = WordCloud(stopwords=stop_words).generate(' '.join(titles))

# Display the generated image:
plt.imshow(title_wordcloud)
plt.axis("off")
plt.show()


# title_wordcloud.words_


# df[df['country']=='India']['title']


titles = df[df['country']=='India']['title']
# Create and generate a word cloud image:
title_wordcloud = WordCloud(stopwords=stop_words).generate(' '.join(titles))

# Display the generated image:
plt.imshow(title_wordcloud)
plt.axis("off")
plt.show()


desc = df[df['country']=='India']['description']
# Create and generate a word cloud image:
desc_wordcloud = WordCloud(stopwords=stop_words).generate(' '.join(desc))

# Display the generated image:
plt.imshow(desc_wordcloud)
plt.axis("off")
plt.show()


titles = df[df['country']=='Canada']['title']
# Create and generate a word cloud image:
title_wordcloud = WordCloud(stopwords=stop_words).generate(' '.join(titles))

# Display the generated image:
plt.imshow(title_wordcloud)
plt.axis("off")
plt.show()


desc = df['description']
# Create and generate a word cloud image:
desc_wordcloud = WordCloud(stopwords=stop_words).generate(' '.join(desc))

# Display the generated image:
plt.imshow(desc_wordcloud)
plt.axis("off")
plt.show()


## words analysis with value counts


def word_analysis(df,columns):
    #lowering the text
    title_process_1=" ".join(df[columns]).lower().split()
    description_process_1=" ".join(df['description']).lower().split()
    #removing punctuation using regex
    title_process_2=[re.sub(r"[^\w\s]",'',t_token) for t_token in title_process_1]
    description_process_2=[re.sub(r"[^\w\s]",'',d_token) for d_token in description_process_1]
    #removing stopwords
    processed_title=[final for final in title_process_2 if final not in stop_words]
    processed_descrp=[final for final in description_process_2 if final not in stop_words]
    
    #using counter function to  get the top 10 most common words
    title_recurrance=Counter(processed_title).most_common(11)[1:]
    description_recuurance=Counter(processed_descrp).most_common(11)[1:]
    
    return title_recurrance,description_recuurance


tile,des=word_analysis(df,'title')


tile

[('love', 170),
 ('2', 129),
 ('man', 81),
 ('christmas', 78),
 ('life', 76),
 ('story', 75),
 ('movie', 73),
 ('world', 70),
 ('little', 64),
 ('one', 63)]

des

[('life', 774),
 ('young', 728),
 ('new', 699),
 ('family', 570),
 ('love', 497),
 ('two', 495),
 ('man', 491),
 ('world', 491),
 ('friends', 466),
 ('woman', 452)]


tile,des=word_analysis(df[df['country']=='India'],'title')


tile

[('2', 16),
 ('ki', 15),
 ('', 14),
 ('aur', 11),
 ('bheem', 10),
 ('love', 10),
 ('mumbai', 10),
 ('dil', 9),
 ('ka', 8),
 ('pyaar', 8)]

des

[('man', 123),
 ('woman', 95),
 ('love', 95),
 ('', 92),
 ('life', 76),
 ('family', 75),
 ('two', 64),
 ('father', 58),
 ('mumbai', 54),
 ('friends', 50)]


tile,des=word_analysis(df[df['country']=='United States'],'title')


tile

[('christmas', 47),
 ('american', 41),
 ('2', 39),
 ('love', 35),
 ('story', 34),
 ('man', 25),
 ('high', 25),
 ('power', 23),
 ('one', 23),
 ('last', 22)]

des

[('new', 287),
 ('life', 274),
 ('documentary', 169),
 ('series', 156),
 ('two', 153),
 ('world', 150),
 ('family', 148),
 ('young', 144),
 ('friends', 134),
 ('school', 127)]


tile,des=word_analysis(df[df['country']=='Canada'],'title')


tile

[('trailer', 10),
 ('boys', 10),
 ('true', 8),
 ('christmas', 8),
 ('monsters', 5),
 ('cop', 5),
 ('day', 4),
 ('', 4),
 ('la', 4),
 ('bad', 4)]

des

[('', 20),
 ('two', 13),
 ('friends', 12),
 ('world', 11),
 ('series', 11),
 ('school', 11),
 ('young', 10),
 ('three', 10),
 ('documentary', 9),
 ('true', 9)]


df['date_added'] = df['date_added'].apply(pd.to_datetime)


last_date = max(df['date_added'])
last_date

Timestamp('2021-09-25 00:00:00')


last_year_data = df[df['date_added'] > last_date - pd.Timedelta(days = 365)]
last_year_data.head()


month_group = last_year_data.groupby(last_year_data['date_added'].map(lambda x: x.month))['show_id'].count()
# Calculating cumulative sum
cumulative_sum = month_group.cumsum()

# Plotting the cumulative sum
sns.lineplot(x=month_group.index, y=cumulative_sum, marker = 'o')
plt.xlabel('Date')
plt.ylabel('Cumulative Count of Contents')
plt.title('Cumulative Count of Contents over the year')
plt.show()


second_last_year_data = df[(df['date_added'] < last_date - pd.Timedelta(days = 365)) & (df['date_added'] > last_date - pd.Timedelta(days = 2*365))]
second_last_year_data.head()


day_group = second_last_year_data.groupby('date_added')['show_id'].count()
# Calculating cumulative sum
cumulative_sum = day_group.cumsum()

# Plotting the cumulative sum
sns.lineplot(x=day_group.index, y=cumulative_sum)
plt.xlabel('Date')
plt.ylabel('Cumulative Count of Contents')
plt.title('Cumulative Count of Contents over the year')
plt.show()


df.columns

Index(['show_id', 'type', 'title', 'director', 'cast', 'country', 'date_added',
       'release_year', 'rating', 'duration', 'listed_in', 'description'],
      dtype='object')


years=df['date_added'].dt.year.value_counts().index.to_list()


years

[2019.0,
 2020.0,
 2018.0,
 2021.0,
 2017.0,
 2016.0,
 2015.0,
 2014.0,
 2011.0,
 2013.0,
 2012.0,
 2009.0,
 2008.0,
 2010.0]


for year in years:
    year_data = df[df['date_added'].dt.year == year]
    month_data = year_data.groupby(year_data['date_added'].dt.month)['show_id'].count()
    # Calculating cumulative sum
    cumulative_sum = month_data.cumsum()

    plt.figure(figsize= (10,8))
    # Plotting the cumulative sum
    sns.lineplot(x=month_data.index, y=cumulative_sum, marker = 'o')
    plt.xlabel('Date')
    plt.ylabel('Cumulative Count of Contents')
    plt.title(f'Cumulative Count of Contents over the year {str(year)[:-2]}')
    plt.show()


def gran_check(df):
    #converting and creating required data,month timeframes
    df['date_added']=pd.to_datetime(df['date_added'])
    df['year']=df['date_added'].dt.year
    df['month']=df['date_added'].dt.month_name()
    
    #year wise grp table
    over_year_count=df.groupby(['year','rating']).size().unstack()
    #filling the null values and change the float to int
    over_year_count=over_year_count.fillna(0).astype(int)
    
    
    #month wise grp table
    month_count_table=df.groupby(['year','month']).size().unstack().fillna(0).astype(int)
    
    
    return over_year_count,month_count_table


over_year_count,month_count_table=gran_check(df)


#plotting the heatamp to visualize the rating-content counts 
plt.figure(figsize=(10,6))
sns.heatmap(over_year_count,annot=True,fmt='g')
plt.title('Content_Heat_map')
plt.xlabel('rating')
plt.ylabel('year')
plt.xticks(rotation=90)
plt.tight_layout()
plt.show()


over_year_count


month_list=df.month.value_counts().index.to_list()
year_list=month_count_table.index.to_list()


month_list

['July',
 'December',
 'September',
 'April',
 'October',
 'August',
 'March',
 'January',
 'June',
 'November',
 'May',
 'February']


academic_year_month_order=['January','February','March','April','May','June'
                           ,'July','August','September','October','November','December']


for year in year_list:
    month_count_table.loc[year].loc[academic_year_month_order].plot(kind='bar')
    plt.xlabel("month_ordered")
    plt.ylabel("Contents")
    plt.xticks(rotation=90)
    plt.title("year_{}".format(year))
    plt.show()


sns.countplot(data = df, y = 'rating', order = df['rating'].value_counts().index.tolist(), palette = 'crest')
plt.show()


df['rating'].value_counts().index.tolist()

['TV-MA',
 'TV-14',
 'TV-PG',
 'R',
 'PG-13',
 'TV-Y7',
 'TV-Y',
 'PG',
 'TV-G',
 'NR',
 'G',
 'TV-Y7-FV',
 'NC-17',
 'UR']


pd.DataFrame(df['rating'].value_counts())

	release_year
count	8807.000000
mean	2014.180198
std	8.819312
min	1925.000000
25%	2013.000000
50%	2017.000000
75%	2019.000000
max	2021.000000

	Country	Count
0	United States	3690
1	India	1046
2	United Kingdom	806
3	Canada	445
4	France	393
...	...	...
118	Ecuador	1
119	Armenia	1
120	Mongolia	1
121	Bahamas	1
122	Montenegro	1

	show_id	type	title	director	cast	country	date_added	release_year	rating	duration	listed_in	description
1957	s1958	Movie	Real Steel	Shawn Levy	Hugh Jackman, Dakota Goyo, Evangeline Lilly, A...	United States, India	2020-09-24	2011	PG-13	127 min	Action & Adventure, Sci-Fi & Fantasy, Sports M...	A struggling fighter-turned-promoter reconnect...
1958	s1959	TV Show	The Chef Show	NaN	Jon Favreau, Roy Choi	United States	2020-09-24	2020	TV-MA	4 Seasons	Docuseries	Writer, director and food enthusiast Jon Favre...
1959	s1960	Movie	Enola Holmes	Harry Bradbeer	Millie Bobby Brown, Henry Cavill, Sam Claflin,...	United Kingdom	2020-09-23	2020	PG-13	124 min	Action & Adventure, Children & Family Movies, ...	While searching for her missing mother, intrep...
1960	s1961	Movie	Kiss the Ground	Joshua Tickell, Rebecca Harrell Tickell	Woody Harrelson	United States	2020-09-22	2020	TV-G	85 min	Documentaries	Science experts and celebrity activists unpack...
1961	s1962	TV Show	The Playbook	NaN	NaN	United States	2020-09-22	2020	TV-MA	1 Season	Docuseries	Coaches with championship résumés share their ...

2024W AML 2103 2 Visualization for AI and ML

Assignment 2

Team Members:¶

Content¶

Inferernce and Results:¶

Q1. How does content differ geographically?¶

Q2. Over time, how has the type of content that is being added, changed?¶

Q3. Which words recurred, among content Titles and Descriptions?¶

pre analysis Q4 ¶

Q4. Over a year, what changes have occurred in the overall content being added?¶

Q5. Which is the most popular rating-category?¶

Dataset Description¶

validation¶

quick_insight¶

Data Validation and Cleansing¶

quick_insight¶

hidden_issue¶

EDA¶

Distribution of content type¶

Directors with highest number of contents directed¶

Most appearing Actors/Actresses¶

Question 1¶

Largest producer Country¶

map_visualization¶

Content Difference according to the countries¶

creating genre based table for the top 10 countries¶

Question 2¶

Type over the time¶

Popular Word Choices¶

Popular word choices for the title of the contents¶

Question 4 pre_analysis¶

year and month analysis¶

Question 4¶

month wise analysis¶

most popular rating¶

	show_id	type	title	director	cast	country	date_added	release_year	rating	duration	listed_in	description
0	s1	Movie	Dick Johnson Is Dead	Kirsten Johnson	NaN	United States	September 25, 2021	2020	PG-13	90 min	Documentaries	As her father nears the end of his life, filmm...
1	s2	TV Show	Blood & Water	NaN	Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban...	South Africa	September 24, 2021	2021	TV-MA	2 Seasons	International TV Shows, TV Dramas, TV Mysteries	After crossing paths at a party, a Cape Town t...
2	s3	TV Show	Ganglands	Julien Leclercq	Sami Bouajila, Tracy Gotoas, Samuel Jouy, Nabi...	NaN	September 24, 2021	2021	TV-MA	1 Season	Crime TV Shows, International TV Shows, TV Act...	To protect his family from a powerful drug lor...

	Columns	Null Count	Null Ratio	Unique Values
0	show_id	0	0.0	8807
1	type	0	0.0	2
2	title	0	0.0	8807
3	director	2634	0.299	4528
4	cast	825	0.094	7692
5	country	831	0.094	748
6	date_added	10	0.001	1714
7	release_year	0	0.0	74
8	rating	4	0.0	17
9	duration	3	0.0	220
10	listed_in	0	0.0	514
11	description	0	0.0	8775

	show_id	type	title	director	cast	country	date_added	release_year	rating	duration	listed_in	description
5541	s5542	Movie	Louis C.K. 2017	Louis C.K.	Louis C.K.	United States	2017-04-04	2017	74 min	NaN	Movies	Louis C.K. muses on religion, eternal love, gi...
5794	s5795	Movie	Louis C.K.: Hilarious	Louis C.K.	Louis C.K.	United States	2016-09-16	2010	84 min	NaN	Movies	Emmy-winning comedy writer Louis C.K. brings h...
5813	s5814	Movie	Louis C.K.: Live at the Comedy Store	Louis C.K.	Louis C.K.	United States	2016-08-15	2015	66 min	NaN	Movies	The comic puts his trademark hilarious/thought...

	Cast Member	Count
0	Anupam Kher	43
1	Shah Rukh Khan	35
2	Julie Tejwani	33
3	Naseeruddin Shah	32
4	Takahiro Sakurai	32
5	Rupa Bhimani	31
6	Akshay Kumar	30
7	Om Puri	30
8	Yuki Kaji	29
9	Paresh Rawal	28
10	Amitabh Bachchan	28
11	Boman Irani	27
12	Rajesh Kava	26
13	Vincent Tong	26
14	Andrea Libman	25
15	Kareena Kapoor	25
16	Samuel L. Jackson	24
17	John Cleese	24
18	Fred Tatasciore	23
19	Jigna Bhardwaj	23

type	country	Movie	TV Show
0	United States	2752	938
1	India	962	84
2	United Kingdom	534	272
3	Canada	319	126
4	France	303	90
5	Japan	119	199
6	Spain	171	61
7	South Korea	61	170
8	Germany	182	44
9	Mexico	111	58

rating	G	NC-17	NR	PG	PG-13	R	TV-14	TV-G	TV-MA	TV-PG	TV-Y	TV-Y7	TV-Y7-FV	UR
country
United States	39	1	43	243	433	660	497	89	1101	304	127	147	2	1
India	0	0	7	7	11	5	572	10	266	144	6	17	1	0
United Kingdom	4	0	12	35	84	145	103	25	253	98	34	12	0	1
Canada	2	1	5	33	32	79	49	17	107	39	45	35	1	0
France	2	1	4	21	35	57	48	6	163	12	21	21	0	2
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
Jamaica	0	0	0	0	0	0	0	0	1	0	0	0	0	0
Slovakia	0	0	0	0	0	0	0	0	1	0	0	0	0	0
Nicaragua	0	0	0	0	0	1	0	0	0	0	0	0	0	0
Ethiopia	0	0	0	0	0	0	0	0	1	0	0	0	0	0
Liechtenstein	0	0	0	0	1	0	0	0	0	0	0	0	0	0

rating	G	NC-17	NR	PG	PG-13	R	TV-14	TV-G	TV-MA	TV-PG	TV-Y	TV-Y7	TV-Y7-FV	UR
year
2008.0	0	0	0	0	0	0	0	0	2	0	0	0	0	0
2009.0	0	0	0	0	0	0	0	0	2	0	0	0	0	0
2010.0	0	0	1	0	0	0	0	0	0	0	0	0	0	0
2011.0	0	0	0	0	0	0	5	0	3	5	0	0	0	0
2012.0	0	0	0	1	0	1	0	0	0	1	0	0	0	0
2013.0	0	0	4	1	0	0	2	1	3	0	0	0	0	0
2014.0	1	0	0	3	0	0	2	1	12	4	1	0	0	0
2015.0	1	0	5	2	2	3	14	5	29	8	7	4	2	0
2016.0	2	1	27	3	6	14	98	9	162	50	10	43	1	0
2017.0	4	1	24	19	26	66	326	23	446	168	35	45	1	1
2018.0	12	0	14	33	53	129	451	36	650	184	40	45	1	0
2019.0	8	1	4	81	135	208	494	40	736	198	54	54	1	2
2020.0	9	0	0	86	122	188	439	61	671	146	102	55	0	0
2021.0	4	0	0	58	146	190	326	44	489	97	57	87	0	0

	rating
TV-MA	3207
TV-14	2160
TV-PG	863
R	799
PG-13	490
TV-Y7	334
TV-Y	307
PG	287
TV-G	220
NR	80
G	41
TV-Y7-FV	6
NC-17	3
UR	3

2024W AML 2103 2 Visualization for AI and ML

Assignment 2

Team Members:¶

Content¶

Inferernce and Results:¶

Q1. How does content differ geographically?¶

Q2. Over time, how has the type of content that is being added, changed?¶

Q3. Which words recurred, among content Titles and Descriptions?¶

pre analysis Q4¶

Q4. Over a year, what changes have occurred in the overall content being added?¶

Q5. Which is the most popular rating-category?¶

Dataset Description¶

validation¶

quick_insight¶

Data Validation and Cleansing¶

quick_insight¶

hidden_issue¶

EDA¶

Distribution of content type¶

Directors with highest number of contents directed¶

Most appearing Actors/Actresses¶

Question 1¶

Largest producer Country¶

map_visualization¶

Content Difference according to the countries¶

creating genre based table for the top 10 countries¶

Question 2¶

Type over the time¶

Popular Word Choices¶

Popular word choices for the title of the contents¶

Question 4 pre_analysis¶

year and month analysis¶

Question 4¶

month wise analysis¶

most popular rating¶

pre analysis Q4 ¶